Deep Learning With an Attention Mechanism for Differentiating the Origin of Brain Metastasis Using MR images.

BACKGROUND: Brain metastasis (BM) is a serious neurological complication of cancer of different origins. The value of deep learning (DL) to identify multiple types of primary origins remains unclear. PURPOSE: To distinguish primary site of BM and identify the best DL models. STUDY TYPE: Retrospective. POPULATION: A total of 449 BM derived from 214 patients (49.5% for female, mean age 58 years) (100 from small cell lung cancer [SCLC], 125 from non-small cell lung cancer [NSCLC], 116 from breast cancer [BC], and 108 from gastrointestinal cancer [GIC]) were included. FIELD STRENGTH/SEQUENCE: A 3-T, T1 turbo spin echo (T1-TSE), T2-TSE, T2FLAIR-TSE, DWI echo-planar imaging (DWI-EPI) and contrast-enhanced T1-TSE (CE T1-TSE). ASSESSMENT: Lesions were divided into training (n = 285, 153 patients), testing (n = 122, 93 patients), and independent testing cohorts (n = 42, 34 patients). Three-dimensional residual network (3D-ResNet), named 3D ResNet6 and 3D ResNet 18, was proposed for identifying the four origins based on single MRI and combined MRI (T1WI + T2-FLAIR + DWI, CE-T1WI + DWI, CE-T1WI + T2WI + DWI). DL model was used to distinguish lung cancer from non-lung cancer; then SCLC vs. NSCLC for lung cancer classification and BC vs. GIC for non-lung cancer classification was performed. A subjective visual analysis was implemented and compared with DL models. Gradient-weighted class activation mapping (Grad-CAM) was used to visualize the model by heatmaps. STATISTICAL TESTS: The area under the receiver operating characteristics curve (AUC) assess each classification performance. RESULTS: 3D ResNet18 with Grad-CAM and AIC showed better performance than 3DResNet6, 3DResNet18 and the radiologist for distinguishing lung cancer from non-lung cancer, SCLC from NSCLC, and BC from GIC. For single MRI sequence, T1WI, DWI, and CE-T1WI performed best for lung cancer vs. non-lung cancer, SCLC vs. NSCLC, and BC vs. GIC classifications. The AUC ranged from 0.675 to 0.876 and from 0.684 to 0.800 regarding the testing and independent testing datasets, respectively. For combined MRI sequences, the combination of CE-T1WI + T2WI + DWI performed better for BC vs. GIC (AUCs of 0.788 and 0.848 on testing and independent testing datasets, respectively), while the combined MRI approach (T1WI + T2-FLAIR + DWI, CE-T1WI + DWI) could not achieve higher AUCs for lung cancer vs. non-lung cancer, SCLC vs. NSCLC. Grad-CAM helped for model visualization by heatmaps that focused on tumor regions. DATA CONCLUSION: DL models may help to distinguish the origins of BM based on MRI data. EVIDENCE LEVEL: 3 TECHNICAL EFFICACY: Stage 2.

Discrimination Between Glioblastoma and Solitary Brain Metastasis Using Conventional MRI and Diffusion-Weighted Imaging Based on a Deep Learning Algorithm.

This study aims to develop and validate a deep learning (DL) model to differentiate glioblastoma from single brain metastasis (BM) using conventional MRI combined with diffusion-weighted imaging (DWI). Preoperative conventional MRI and DWI of 202 patients with solitary brain tumor (104 glioblastoma and 98 BM) were retrospectively obtained between February 2016 and September 2022. The data were divided into training and validation sets in a 7:3 ratio. An additional 32 patients (19 glioblastoma and 13 BM) from a different hospital were considered testing set. Single-MRI-sequence DL models were developed using the 3D residual network-18 architecture in tumoral (T model) and tumoral + peritumoral regions (T&P model). Furthermore, the combination model based on conventional MRI and DWI was developed. The area under the receiver operating characteristic curve (AUC) was used to assess the classification performance. The attention area of the model was visualized as a heatmap by gradient-weighted class activation mapping technique. For the single-MRI-sequence DL model, the T2WI sequence achieved the highest AUC in the validation set with either T models (0.889) or T&P models (0.934). In the combination models of the T&P model, the model of DWI combined with T2WI and contrast-enhanced T1WI showed increased AUC of 0.949 and 0.930 compared with that of single-MRI sequences in the validation set, respectively. And the highest AUC (0.956) was achieved by combined contrast-enhanced T1WI, T2WI, and DWI. In the heatmap, the central region of the tumoral was hotter and received more attention than other areas and was more important for differentiating glioblastoma from BM. A conventional MRI-based DL model could differentiate glioblastoma from solitary BM, and the combination models improved classification performance.

Limb Remote Ischemic Preconditioning Applied During Sevoflurane Anesthesia Does Not Protect the Lungs in Patients Undergoing Adult Heart Valve Surgery.

BACKGROUND: Two consistent overall cell protective preconditioning treatments should provide more protection. We hypothesized that limb remote ischemic preconditioning (RIPC, second preconditioning stimulus) applied during sevoflurane inhalation (first preconditioning stimulus) would provide more protection to the lungs of patients undergoing adult heart valve surgery. METHODS: In this randomized, placebo-controlled, double-blind trial, 50 patients were assigned to the RIPC group or the placebo group (1:1). Patients in the RIPC group received three 5-min cycles of 300 mmHg cuff inflation/deflation of the left-side lower limb before aortic cross-clamping. Anesthesia consisted of opioids and propofol for induction and sevoflurane for maintenance. The primary end point was comparison of the postoperative arterial-alveolar oxygen tension ratio (a/A ratio) between groups. Secondary end points included comparisons of pulmonary variables, postoperative morbidity and mortality and regional and systemic inflammatory cytokines between groups. RESULTS: In the RIPC group, the a/A ratio and other pulmonary variables exhibited no significant differences throughout the study period compared with the placebo group. No significant differences in either plasma or bronchoalveolar lavage levels of TNF- α were noted between the groups at 10 min after anesthetic induction and 1 h after cross-clamp release. The percentage of neutrophils at 12 h postoperation was significantly increased in the RIPC group compared with the placebo group (91.34±0.00 vs. 89.42±0.10, P = 0.023). CONCLUSIONS: Limb RIPC applied during sevoflurane anesthesia did not provide additional significant pulmonary protection following adult valvular cardiac surgery.

Outcomes of Totally Endoscopic Beating-Heart Tricuspid Repair in Redo Cardiac Surgery.

BACKGROUND: The role of endoscopic surgery in treating late severe tricuspid regurgitation after cardiac surgery has not been well investigated. The aim of this study was to evaluate the outcomes of a combination of a beating-heart, minimally invasive approach and a leaflet-augmentation technique in treating tricuspid regurgitation after cardiac surgery. METHOD: This was a retrospective cohort study. From January 2015 to July 2018, patients undergoing reoperative tricuspid valve repair with a totally endoscopic approach were enrolled. Procedures were performed on beating hearts with normothermic cardiopulmonary bypass (CPB). RESULTS: A total of 43 adults (mean age 53.4±11.4 yr; 9 men) met the inclusion criteria. The interval between prior cardiac surgery and current tricuspid repair was 17.6±6.5 years. Ten (10) patients had previous tricuspid repair and concomitant previous cardiac surgery. In the current endoscopic approach, tricuspid repair techniques included 38 leaflet augmentations, 38 annular ring placements, five artificial chordae, one cleft closure, five commissure recreations, and eight papillary muscle relaxations. Mean CPB time, median ventilation time, and median hospital stay were 128.5±54.2 minutes, 20.5 hours (range, 6-436 hrs), and 7 days (range, 4-56 d), respectively. There were only three in-hospital deaths and no follow-up mortality. The regurgitant jet area was decreased from 21.5±12.1 cm2 preoperatively to 2.4±2.2 cm2 postoperatively (p<0.001). In patients with previous tricuspid repair, although the technique of valvuloplasty seems more complex, CPB time, procedure time and hospital stay were not longer than in patients who did not have previous tricuspid repair. CONCLUSIONS: Beating-heart, video-assisted, minimal access tricuspid repair after previous cardiac surgery is feasible, reproducible, and associated with low mortality, even in patients who have had previous tricuspid repair.

Cho/Cr ratio at MR spectroscopy as a biomarker for cellular proliferation activity and prognosis in glioma: correlation with the expression of minichromosome maintenance protein 2.

BACKGROUND: Magnetic resonance (MR) spectroscopy (1H-MRS) has been demonstrated to be useful in grading glioma, but the utility in assessing cellular proliferation activity and prognosis correlated with the expression of minichromosome maintenance protein 2 (MCM2) has not been reported. PURPOSE: To explore the correlation between proton MR spectroscopy parameters (including choline [Cho]/creatine [Cr], N-acetyl aspartate [NAA]/Cr, and Cho/NAA ratios) and the expression of MCM2 and to further evaluate whether 1H-MRS can predict cell proliferative activity and provide prognostic information in high-grade gliomas (HGGs). MATERIAL AND METHODS: Forty-three patients with histopathologically confirmed gliomas were involved in this study. All patients underwent 1H-MRS examination before surgery. Proliferative activity of gliomas was evaluated by MCM2 labeling index (LI). Pearson correlation analysis and empiric receiver operating characteristic (ROC) curves were performed. The Kaplan-Meier method and Cox regression were used for survival analysis. RESULTS: Significant correlation was observed between the Cho/Cr ratio and MCM2 LI ( r = 0.522, P < 0.01); however, there was no correlation between MCM2 LI and the Cho/NAA or NAA/Cr ratios ( r = 0.295, P = 0.55 and r = -0.042, P = 0.788, respectively). According to ROC analysis, MCM2 LI of 50% and Cho/Cr ratio of 2.68 represented the optimized cut-off values, respectively, to distinguish longer or shorter survival than 15 months in HGGs patients. Multivariate analysis revealed that both the Cho/Cr ratio and MCM2 expression were independent prognostic markers. CONCLUSION: Cho/Cr ratio has a potential in predicting the expression of MCM2 and can evaluate cell proliferative activity noninvasively. Both the Cho/Cr ratio and MCM2 expression are independent prognostic markers in patients with HGGs.

Measuring the lactate-to-creatine ratio via 1H NMR spectroscopy can be used to noninvasively evaluate apoptosis in glioma cells after X-ray irradiation.

BACKGROUND: Radiotherapy is among the commonly applied treatment options for glioma, which is one of the most common types of primary brain tumor. To evaluate the effect of radiotherapy noninvasively, it is vital for oncologists to monitor the effects of X-ray irradiation on glioma cells. Preliminary research had showed that PKC-ι expression correlates with tumor cell apoptosis induced by X-ray irradiation. It is also believed that the lactate-to-creatine (Lac/Cr) ratio can be used as a biomarker to evaluate apoptosis in glioma cells after X-ray irradiation. In this study, we evaluated the relationships between the Lac/Cr ratio, apoptotic rate, and protein kinase C iota (PKC-ι) expression in glioma cells. METHODS: Cells of the glioma cell lines C6 and U251 were randomly divided into 4 groups, with every group exposed to X-ray irradiation at 0, 1, 5, 10 and 15 Gy. Single cell gel electrophoresis (SCGE) was conducted to evaluate the DNA damage. Flow cytometry was performed to measure the cell cycle blockage and apoptotic rates. Western blot analysis was used to detect the phosphorylated PKC-ι (p-PKC-ι) level. 1H NMR spectroscopy was employed to determine the Lac/Cr ratio. RESULTS: The DNA damage increased in a radiation dose-dependent manner (p < 0.05). With the increase in X-ray irradiation, the apoptotic rate also increased (C6, p < 0.01; U251, p < 0.05), and the p-PKC-ι level decreased (C6, p < 0.01; U251, p < 0.05). The p-PKC-ι level negatively correlated with apoptosis, whereas the Lac/Cr ratio positively correlated with the p-PKC-ι level. CONCLUSION: The Lac/Cr ratio decreases with an increase in X-ray irradiation and thus can be used as a biomarker to reflect the effects of X-ray irradiation in glioma cells.

Bilateral subclavian vein sheaths for superior vena cava drainage during thoracoscopic repair of atrial septal defects: cosmetic outcomes, safety and effectiveness.

OBJECTIVE: To investigate the cosmetic outcomes, safety and effectiveness of using bilateral subclavian vein sheaths for superior vena cava drainage during thoracoscopic repair of atrial septal defects. METHODS: Sixty-one consecutive adults scheduled for thoracoscopic repair of atrial septal defects between July 2012 and June 2013 were randomized into two groups: one group underwent placement of a 16 Fr percutaneous superior vena cava cannula (n = 30) and the other group underwent placement of bilateral 8 Fr subclavian vein sheaths (n = 31) for superior vena cava drainage during peripheral cardiopulmonary bypass. The perioperative data, central venous pressure during cardiopulmonary bypass, complications and the patient satisfaction scale scores for the incisions were compared between the two groups. RESULTS: The theoretical cardiopulmonary bypass flow rate was reached without complications in all patients. The average central venous pressure during cardiopulmonary bypass was not significantly different between the two groups [(6.9 ± 3.1) mmHg vs. (7.0 ± 3.5) mmHg, p=0.92]. The patient satisfaction scale scores for the incisions were significantly higher in the patients who underwent placement of bilateral subclavian vein sheaths than in the patients who underwent placement of a percutaneous superior vena cava cannula [(2.81 ± 0.75) vs. (2.07 ± 0.74), p<0.001]. CONCLUSIONS: Placement of bilateral subclavian vein sheaths is a safe and effective alternative to placement of a percutaneous superior vena cava cannula for superior vena cava drainage during thoracoscopic repair of atrial septal defects and results in greater patient satisfaction with the cosmetic outcome.

Assessment of the intrinsic radiosensitivity of glioma cells and monitoring of metabolite ratio changes after irradiation by 14.7-T high-resolution ¹H MRS.

Gliomas are the most common type of primary brain tumor. Radiation therapy (RT) is the primary adjuvant treatment to eliminate residual tumor tissue after surgery. However, the current RT guided by conventional imaging is unsatisfactory. A fundamental question is whether it is possible to further enhance the effectiveness and efficiency of RT based on individual radiosensitivity. In this research, to probe the correlation between radiosensitivity and the metabolite characteristics of glioma cells in vitro, a perchloric acid (PCA) extracting method was used to obtain water-soluble metabolites [such as N-acetylaspartate (NAA), choline (Cho), creatine (Cr) and succinate (Suc)]. Spectral patterns from these processed water-soluble metabolite samples were acquired by in vitro 14.7-T high-resolution ¹H MRS. Survival fraction analysis was performed to test the intrinsic radiosensitivity of glioma cell lines. Good ¹H MRS of PCA extracts from glioma cells was obtained. The radiosensitivity of glioma cells correlated positively with the Cho/Cr and Cho/NAA ratios, but negatively with the Suc/Cr ratio. Irradiation of the C6 cell line at different X-ray dosages led to changes in metabolite ratios and apoptotic rates. A plateau phase of metabolite ratio change and a decrease in apoptotic rate were found in the C6 cell line. We conclude that in vitro high-resolution ¹H MRS possesses the sensitivity required to detect subtle biochemical changes at the cellular level. ¹H MRS may aid in the assessment of the individual radiosensitivity of brain tumors, which is pivotal in the identification of the biological target volume.

Application of magnetic resonance spectroscopy in patients with alternating hemiplegia of childhood: findings on metabolic dysfunctions.

OBJECTIVE: Alternating hemiplegia of childhood (AHC) is a rare neurodevelopmental syndrome of uncertain etiology. Although the use of magnetic resonance spectroscopy (MRS) for the study of neurologic diseases has grown rapidly over the past decade, its use for AHC patients is quite new. This study was aimed at investigating changes of brain metabolites in patients with alternating hemiplegia of childhood (AHC) during the hemiplegic ictal phases and interictal phases by proton magnetic resonance spectroscopy ((1)H-MRS). METHODS: (1)H-MRS was used in AHC patients during the hemiplegic ictal phases and interictal phases to evaluate functional activity in certain brain regions. A total of 10 unmedicated, healthy volunteers served as controls. RESULTS: N-acetylaspartate (NAA)/Creatine(Cr) ratio of the frontal lobes, basal ganglia, and temporal lobes in contralateral hemiplegic hemisphere of AHC patients during the ictal phases was significantly lower than that in AHC patients during interictal phases and control subjects. Significantly increased choline-containing compounds (Cho)/Cr were obtained in corresponding regions. CONCLUSIONS: These findings suggest neuronal metabolic dysfunctions in frontal lobes, temporal lobes and basal ganglia in AHC patients during ictal phases that perhaps are involved in the pathogenesis of AHC.

Superior vena cava drainage during thoracoscopic cardiac surgery: bilateral internal jugular vein sheaths versus one percutaneous superior vena cava cannula.

OBJECTIVE: To evaluate bilateral internal jugular vein sheaths as a replacement of one percutaneous superior vena cava cannula for superior vena cava drainage during thoracoscopic cardiac surgery. DESIGN: A prospective and randomized study. SETTING: Single cardiovascular institute. PARTICIPANTS: Adults undergoing thoracoscopic cardiac surgery. INTERVENTIONS: Patients were randomized into a percutaneous superior vena cava cannula group and a bilateral internal jugular vein sheaths group. The superior vena cava drainage for cardiopulmonary bypass was performed with one percutaneous superior vena cava cannula (14-18 Fr) or the bilateral internal jugular vein sheaths (8 Fr). MEASUREMENTS AND MAIN RESULTS: Both interventions reached theoretic flow rate in all patients. In patients weighing<50 kg (n=38) and 50-70 kg (n=64), the average central venous pressure values during cardiopulmonary bypass of both groups showed no significant differences. The patients weighing>70 kg (n=15) in the bilateral internal jugular vein sheaths group had a normal average central venous pressure value, but it was significantly higher than that of percutaneous superior vena cava cannula group ([10.5±3.1] mmHg vs. [4.5±4.4] mmHg, p=0.013). The patient satisfaction scale scores for the cervical incisions were significantly higher in the bilateral internal jugular vein sheaths group than in the percutaneous superior vena cava cannula group ([2.6±0.9] vs. [2.1±0.8], p=0.002). CONCLUSIONS: The bilateral internal jugular vein sheaths were a feasible and effective option to replace one percutaneous superior vena cava cannula during thoracoscopic cardiac surgery, with better patient satisfaction.

Preoperatively Predicting Ki67 Expression in Pituitary Adenomas Using Deep Segmentation Network and Radiomics Analysis Based on Multiparameter MRI.

RATIONALE AND OBJECTIVES: Ki67 proliferation index is associated with more aggressive tumor behavior and recurrence of pituitary adenomas (PAs). Recently, radiomics and deep learning have been introduced into the study of pituitary tumors. The present study aimed to investigate the feasibility of predicting the Ki67 proliferation index of PAs using the deep segmentation network and radiomics analysis based on multiparameter MRI. MATERIALS AND METHODS: First, the cfVB-Net autosegmentation model was trained; subsequently, its performance was evaluated in terms of the dice similarity coefficient (DSC). In the present study, 1214 patients were classified into the high Ki67 expression group (HG) and the low Ki67 expression group (LG). Analyses of three classification models based on radiomics features were performed to distinguish HG from LG. Clinical factors, imaging features, and Radscores were collectively used to create a nomogram in order to effectively predict Ki67 expression. RESULTS: The cfVB-Net segmentation model demonstrated good performance (DSC: 0.723-0.930). Overall, 18, 15, and 11 optimal features in contrast-enhanced (CE) T1WI, T1WI, and T2WI were obtained for differentiating between HG and LG, respectively. Notably, the best results were presented in the bagging decision tree when CE T1WI and T1WI were combined (area under the receiver operating characteristic curve: training set, 0.927; validation set, 0.831; and independent testing set, 0.825). In the nomogram, age, Hardy' grade, and Radscores were identified as risk predictors of high Ki67 expression. CONCLUSION: The deep segmentation network and radiomics analysis based on multiparameter MRI exhibited good performance and clinical application value in predicting the expression of Ki67 in PAs.

Identification of Prolactinoma in Pituitary Neuroendocrine Tumors Using Radiomics Analysis Based on Multiparameter MRI.

This study aims to investigate the feasibility of preoperatively predicting histological subtypes of pituitary neuroendocrine tumors (PitNETs) using machine learning and radiomics based on multiparameter MRI. Patients with PitNETs from January 2016 to May 2022 were retrospectively enrolled from four medical centers. A cfVB-Net network was used to automatically segment PitNET multiparameter MRI. Radiomics features were extracted from the MRI, and the radiomics score (Radscore) of each patient was calculated. To predict histological subtypes, the Gaussian process (GP) machine learning classifier based on radiomics features was performed. Multi-classification (six-class histological subtype) and binary classification (PRL vs. non-PRL) GP model was constructed. Then, a clinical-radiomics nomogram combining clinical factors and Radscores was constructed using the multivariate logistic regression analysis. The performance of the models was evaluated using receiver operating characteristic (ROC) curves. The PitNET auto-segmentation model eventually achieved the mean Dice similarity coefficient of 0.888 in 1206 patients (mean age 49.3 ± SD years, 52% female). In the multi-classification model, the GP of T2WI got the best area under the ROC curve (AUC), with 0.791, 0.801, and 0.711 in the training, validation, and external testing set, respectively. In the binary classification model, the GP of T2WI combined with CE T1WI demonstrated good performance, with AUC of 0.936, 0.882, and 0.791 in training, validation, and external testing sets, respectively. In the clinical-radiomics nomogram, Radscores and Hardy' grade were identified as predictors for PRL expression. Machine learning and radiomics analysis based on multiparameter MRI exhibited high efficiency and clinical application value in predicting the PitNET histological subtypes.

Multi-parameter MRI-Based Machine Learning Model to Evaluate the Efficacy of STA-MCA Bypass Surgery for Moyamoya Disease: A Pilot Study.

Superficial temporal artery-middle cerebral artery (STA-MCA) bypass surgery represents the primary treatment for Moyamoya disease (MMD), with its efficacy contingent upon collateral vessel development. This study aimed to develop and validate a machine learning (ML) model for the non-invasive assessment of STA-MCA bypass surgery efficacy in MMD. This study enrolled 118 MMD patients undergoing STA-MCA bypass surgery. Clinical features were screened to construct a clinical model. MRI features were extracted from the middle cerebral artery supply area using 3D Slicer and employed to build five ML models using logistic regression algorithm. The combined model was developed by integrating the radiomics score (Rad-score) with the clinical features. Model performance validation was conducted using ROC curves. Platelet count (PLT) was identified as a significant clinical feature for constructing the clinical model. A total of 3404 features (851 × 4) were extracted, and 15 optimal features were selected from each MRI sequence as predictive factors. Multivariable logistic regression identified PLT and Rad-score as independent parameters used for constructing the combined model. In the testing set, the AUC of the T1WI ML model [0.84 (95% CI, 0.70-0.97)] was higher than that of the clinical model [0.66 (95% CI, 0.46-0.86)] and the combined model [0.80 (95% CI, 0.66-0.95)]. The T1WI ML model can be used to assess the postoperative efficacy of STA-MCA bypass surgery for MMD.

Characterization of prostatic cancer lesion and gleason grade using a continuous-time random-walk diffusion model at high b-values.

Background: Distinguishing between prostatic cancer (PCa) and chronic prostatitis (CP) is sometimes challenging, and Gleason grading is strongly associated with prognosis in PCa. The continuous-time random-walk diffusion (CTRW) model has shown potential in distinguishing between PCa and CP as well as predicting Gleason grading. Purpose: This study aimed to quantify the CTRW parameters (α, ß & Dm) and apparent diffusion coefficient (ADC) of PCa and CP tissues; and then assess the diagnostic value of CTRW and ADC parameters in differentiating CP from PCa and low-grade PCa from high-grade PCa lesions. Study type: Retrospective (retrospective analysis using prospective designed data). Population: Thirty-one PCa patients undergoing prostatectomy (mean age 74 years, range 64-91 years), and thirty CP patients undergoing prostate needle biopsies (mean age 68 years, range 46-79 years). Field strength/Sequence: MRI scans on a 3.0T scanner (uMR790, United Imaging Healthcare, Shanghai, China). DWI were acquired with 12 b-values (0, 50, 100, 150, 200, 500, 800, 1200, 1500, 2000, 2500, 3000 s/mm2). Assessment: CTRW parameters and ADC were quantified in PCa and CP lesions. Statistical tests: The Mann-Whitney U test was used to evaluate the differences in CTRW parameters and ADC between PCa and CP, high-grade PCa, and low-grade PCa. Spearman's correlation of the pathologic grading group (GG) with CTRW parameters and ADC was evaluated. The usefulness of CTRW parameters, ADC, and their combinations (Dm, α and ß; Dm, α, ß, and ADC) to differentiate PCa from CP and high-grade PCa from low-grade PCa was determined by logistic regression and receiver operating characteristic curve (ROC) analysis. Delong test was used to compare the differences among AUCs. Results: Significant differences were found for the CTRW parameters (α, Dm) between CP and PCa (all P<0.001), high-grade PCa, and low-grade PCa (α:P=0.024, Dm:P=0.021). GG is correlated with certain CTRW parameters and ADC(α:P<0.001,r=-0.795; Dm:P<0.001,r=-0.762;ADC:P<0.001,r=-0.790). Moreover, CTRW parameters (α, ß, Dm) combined with ADC showed the best diagnostic efficacy for distinguishing between PCa and CP as well as predicting Gleason grading. The differences among AUCs of ADC, CTRW parameters and their combinations were not statistically significant (P=0.051-0.526). Conclusion: CTRW parameters α and Dm, as well as their combination were beneficial to distinguish between CA and PCa, low-grade PCa and high-grade PCa lesions, and CTRW parameters and ADC had comparable diagnostic performance.

Deep Learning Radiomics Model of Contrast-Enhanced CT for Differentiating the Primary Source of Liver Metastases.

RATIONALE AND OBJECTIVES: To develop and validate a deep learning radiomics (DLR) model based on contrast-enhanced computed tomography (CT) to identify the primary source of liver metastases. MATERIALS AND METHODS: In total, 657 liver metastatic lesions, including breast cancer (BC), lung cancer (LC), colorectal cancer (CRC), gastric cancer (GC), and pancreatic cancer (PC), from 428 patients were collected at three clinical centers from January 2018 to October 2023 series. The lesions were randomly assigned to the training and validation sets in a 7:3 ratio. An additional 112 lesions from 61 patients at another clinical center served as an external test set. A DLR model based on contrast-enhanced CT of the liver was developed to distinguish the five pathological types of liver metastases. Stepwise classification was performed to improve the classification efficiency of the model. Lesions were first classified as digestive tract cancer (DTC) and non-digestive tract cancer (non-DTC). DTCs were divided into CRC, GC, and PC and non-DTCs were divided into LC and BC. To verify the feasibility of the DLR model, we trained classical machine learning (ML) models as comparison models. Model performance was evaluated using accuracy (ACC) and area under the receiver operating characteristic curve (AUC). RESULTS: The classification model constructed by the DLR algorithm showed excellent performance in the classification task compared to ML models. Among the five categories task, highest ACC and average AUC were achieved at 0.563 and 0.796 in the validation set, respectively. In the DTC and non-DTC and the LC and BC classification tasks, AUC was achieved at 0.907 and 0.809 and ACC was achieved at 0.843 and 0.772, respectively. In the CRC, GC, and PC classification task, ACC and average AUC were the highest, at 0.714 and 0.811, respectively. CONCLUSION: The DLR model is an effective method for identifying the primary source of liver metastases.

Identifying Pathological Subtypes of Brain Metastasis from Lung Cancer Using MRI-Based Deep Learning Approach: A Multicenter Study.

The aim of this study was to investigate the feasibility of deep learning (DL) based on multiparametric MRI to differentiate the pathological subtypes of brain metastasis (BM) in lung cancer patients. This retrospective analysis collected 246 patients (456 BMs) from five medical centers from July 2016 to June 2022. The BMs were from small-cell lung cancer (SCLC, n = 230) and non-small-cell lung cancer (NSCLC, n = 226; 119 adenocarcinoma and 107 squamous cell carcinoma). Patients from four medical centers were assigned to training set and internal validation set with a ratio of 4:1, and we selected another medical center as an external test set. An attention-guided residual fusion network (ARFN) model for T1WI, T2WI, T2-FLAIR, DWI, and contrast-enhanced T1WI based on the ResNet-18 basic network was developed. The area under the receiver operating characteristic curve (AUC) was used to assess the classification performance. Compared with models based on five single-sequence and other combinations, a multiparametric MRI model based on five sequences had higher specificity in distinguishing BMs from different types of lung cancer. In the internal validation and external test sets, AUCs of the model for the classification of SCLC and NSCLC brain metastasis were 0.796 and 0.751, respectively; in terms of differentiating adenocarcinoma from squamous cell carcinoma BMs, the AUC values of the prediction models combining the five sequences were 0.771 and 0.738, respectively. DL together with multiparametric MRI has discriminatory feasibility in identifying pathology type of BM from lung cancer.

Diffusion tensor imaging of normal-appearing white matter in unilateral cerebral arterial occlusive disease.

PURPOSE: To use MR with diffusion tensor imaging (DTI) and conventional and high b value to assess diffusion changes in normal-appearing white matter (NAWM) in patients with unilateral, severe stenosis, or occlusion of the middle cerebral artery (MCA). MATERIALS AND METHODS: In total, 28 patients with NAWM and unilateral, severe stenosis, or occlusion of the MCA underwent DTI with b values 1000 and 2200 s/mm(2) at 3.0T MR. Fractional anisotropy (FA), apparent diffusion coefficient (ADC), radial diffusivity (eigenvalues λ1 , λ2 ), and axial diffusivity (eigenvalue λ3 ) were measured for the ipsilateral and contralateral corona radiata. RESULTS: Mean FA was significantly lower for the ipsilateral than contralateral corona radiata with high b value, 2200 s/mm(2) , and ipsilateral corona radiata with conventional low b value, 1000 s/mm(2) (all P < 0.01). Mean ADC, λ1 , λ2 , and λ3 were significantly higher for the ipsilateral than contralateral corona radiata with high b value (all P < 0.05) but not for ipsilateral than contralateral corona radiata with low b value (P > 0.05). CONCLUSION: DTI with a high b value detects diffusion changes in NAWM in patients with unilateral, severe stenosis, or occlusion of the MCA not seen with conventional b value or conventional MRI contrasts.

Using the metabolite alterations monitoring the AEG-1 expression level and cell biological behaviour of U251 cell in vitro.

Astrocyte elevated gene-1 (AEG-1) is an important oncogene that overexpresses in gliomas and plays a vital role in their occurrence and progression. However, few reports have shown which biomarkers could reflect the level of AEG-1 expression in vivo so far. In recent years, intracellular metabolites monitored by proton magnetic resonance spectroscopy (1H MRS) as non-invasive imaging biomarkers have been applied to the precise diagnosis and therapy feedback of gliomas. Therefore, understanding the correlation between 1H MRS metabolites and AEG-1 gene expression in U251 cells may help to identify relevant biomarkers. This study constructed three monoclonal AEG-1-knockout U251 cell lines using the clustered regularly interspaced short palindromic repeat (CRISPR) /Cas9 technique and evaluated the biological behaviors and metabolite ratios of these cell lines. With the decline in AEG-1 expression, the apoptosis rate of the AEG-1-knockout cell lines increased. At the same time, the metastatic capacities decreased, and the relative contents of total choline (tCho) and lactate (Lac) were also reduced. In conclusion, deviations in AEG-1 expression influence the apoptosis rate and metastasis capacity of U251 cells, which the 1H MRS metabolite ratio could monitor. The tCho/creatinine(Cr) and Lac/Cr ratios positively correlated with the AEG-1 expression and malignant cell behavior. This study may provide potential biomarkers for accurate preoperative diagnosis and future AEG-1-targeting treatment evaluation of gliomas in vivo.

Differentiation of Lung Metastases Originated From Different Primary Tumors Using Radiomics Features Based on CT Imaging.

RATIONALE AND OBJECTIVES: To explore the feasibility of differentiating three predominant metastatic tumor types using lung computed tomography (CT) radiomics features based on supervised machine learning. MATERIALS AND METHODS: This retrospective analysis included 252 lung metastases (LM) (from 78 patients), which were divided into the training (n = 176) and test (n = 76) cohort randomly. The metastases originated from colorectal cancer (n = 97), breast cancer (n = 87), and renal carcinoma (n = 68). An additional 77 LM (from 35 patients) were used for external validation. All radiomics features were extracted from lung CT using an open-source software called 3D slicer. The least absolute shrinkage and selection operator (LASSO) method selected the optimal radiomics features to build the model. Random forest and support vector machine (SVM) were selected to build three-class and two-class models. The performance of the classification model was evaluated with the area under the receiver operating characteristic curve (AUC) by two strategies: one-versus-rest and one-versus-one. RESULTS: Eight hundred and fifty-one quantitative radiomics features were extracted from lung CT. By LASSO, 23 optimal features were extracted in three-class, and 25, 29, and 35 features in two-class for differentiating every two of three LM (colorectal cancer vs. renal carcinoma, colorectal cancer vs. breast cancer, and breast cancer vs. renal carcinoma, respectively). The AUCs of the three-class model were 0.83 for colorectal cancer, 0.79 for breast cancer, and 0.91 for renal carcinoma in the test cohort. In the external validation cohort, the AUCs were 0.77, 0.83, and 0.81, respectively. Swarmplot shows the distribution of radiomics features among three different LM types. In the two-class model, high accuracy and AUC were obtained by SVM. The AUC of discriminating colorectal cancer LM from renal carcinoma LM was 0.84, and breast cancer LM from colorectal cancer LM and renal carcinoma LM were 0.80 and 0.94, respectively. The AUCs were 0.77, 0.78, and 0.84 in the external validation cohort. CONCLUSION: Quantitative radiomics features based on Lung CT exhibited good discriminative performance in LM of primary colorectal cancer, breast cancer, and renal carcinoma.

Distinguishing Parotid Polymorphic Adenoma and Warthin Tumor Based on the CT Radiomics Nomogram: A Multicenter Study.

RATIONALE AND OBJECTIVES: To develop, validate, and test a comprehensive radiomics prediction model to distinguish parotid polymorphic adenomas (PAs) and warthin tumors (WTs) using clinical data and enhanced computed tomography (CT) from a multicenter cohort. MATERIALS AND METHODS: A total of 267 patients with PAs (n =172) or WTs (n = 95) from two hospitals were randomly divided into training (n =188) and validation (n =79) datasets. Radiomics features were extracted from the enhanced CT (arterial phase) followed by dimensionality reduction. Clinical and CT features were combined to establish a prediction model. A radiomics nomogram was constructed by combining RadScore and clinical factors. Moreover, an independent dataset of 31 patients from a third hospital was employed to test the model. Thus, the performance of the nomogram, radiomics signature, and clinical models was evaluated on the training, validation, and the independent testing datasets. Receiver operating characteristic (ROC) curves were used to compare the performance, and decision curve analysis (DCA) was used to evaluate the clinical effectiveness of the model. RESULTS: A total of 15 radiomics features were selected from CT data as the imaging markers to generate RadScores, and demographics or clinical data like age, sex, and smoking factors combined with RadScores were used to distinguish PAs and WTs based on multivariate logistic regression analyses. The results showed that radiomics nomograms combining clinical factors and RadScores provided satisfactory predictive values for distinguishing PAs from WTs, with areas under ROC curves (AUC) of 0.979, 0.922, and 0.903 for the training, validation, and the independent testing datasets, respectively. Decision curve analysis revealed that the radiomics nomogram outperformed the clinical factor models in terms of accuracy and effectiveness. CONCLUSION: CT-based radiomics nomograms combining RadScores and clinical factors can be used to identify PAs and WTs, which may help tumor management by clinicians.